We will continue ongoing studies on the development of immunotoxins (ITs) for in vitro and in vivo use. We will concentrate on the use of ricin A chain linked to intact or Fab fragments of cell-reactive antibodies. Two B cell tumor models will be used; the human Daudi cells and the murine BCL1 tumor. Several linkages between antibody and A chain will be compared for efficacy and in vivo stability. Fab-A immunotoxins and the intact antibody-A-ITs will be studied in vivo for their rate of clearance, biodistribution, homing to target cells, damage to normal tissue and uptake by cells of the reticuloendothelial system. In addition, we will use the "piggyback" approach of cell-reactive A-IT and anti-antibody linked to ricin B chain (B-IT) or an anti-A chain linked to ricin B chain in an attempt to further potentiate the killing by the A-ITs. We will also study the in vitro properties of these B-ITs and compare the piggyback approach with univalent vs divalent antibodies for their ability to delete tumor cells and not damage normal marrow CFU. Clonogenic (Daudi) and adoptive transfer assays (BCL1) will be used as a readout system. The effect of lysosomotropic agents on the piggyback system (tumor cells and bone marrow) will also be studied. Finally, the long-term goal will be to clone and express ricin A and B chains entirely by recombinant DNA technology and eventually to prepare ITs in the same manner. We will mutagenize the A and B chain genes to obtain molecules with more favorable biological properties such as reduced antigenecity, loss of galactose-binding sites on the B chain, and lack of oligosaccharides. The insights gained from this proposal should be applicable to immunotoxins constructed with any type of monoclonal antibody, even though our own studies will be carried out in two defined tumor model systems.